Question from the Machine learning - Fundamentals test

Variance and bias are both linked together : increasing one means decreasing the other. One needs to find a good equilibrium between the two.

The dilemma is that it’s hard to find a way to both minimize bias and variance but we need the two to be as small as possible

If variance and bias are too low, the model will be unable to adapt to new data sets : one need to find an efficiency-adaptability equilibrium

The dilemma consists in finding an efficient way to either compute variance or bias : it can be very hard depending on the model’s structure